Ancient and Recent Patterns of Geographic Speciation in the Oyster

Proc. Nati. Acad. Sci. USA Vol. 91, pp. 4599-4603, May 1994 Evolution Ancient and recent patterns of geographic speciation in the oyster mushroom Pleurotus revealed by phylogenetic analysis of ribosomal DNA sequences (biogeogaphy/specis concept/vcarau) RYTAS VILGALYS AND BAO LIN SUN Department of Botany, Duke University, Durham, NC 27708-0338 Communicated by John C. Avise, February 7, 1994 (receivedfor review December 9, 1993)

ABSTRACT Evidence from molecular systematic studies Saprobic basidiomycetes such as Pleurotus present an ex- suggests that many mushroom species may be quite ancient. cellent system for analysis of speciation in fungi. Many Gene phylogenies were developed to examine the relationship species may be grown and fruited in pure culture, permitting between reproductive isolation, genetic divergence, and bioge- analysis of their mating relationships and other features. ography in oyster mushrooms (Plurotus). Sequence data were Phylogenetic analyses using molecular sequence data partic- obtained for two regions ofDNA from populations belonging to ularly show much promise for resolving phylogenetic rela- eight intersterility groups (biological species). Phylogenetic tionships and understanding speciation for many problematic analysis ofsequences from the 5' portion ofthe nuclear encoded species complexes in basidiomycetes (4-6). large subunit rDNA demonstrates an ancient origin for four In this study we analyzed (i) mating compatibility relation- intersterility groups of broad geographic distribution (world- ships among collections from different parts ofthe world that wide), with a more recent radiation of several intersterility reveal at least eight intersterility groups in Pleurotus and (ii) groups that are restricted to the Northern Hemisphere. An gene phylogenies for two different regions of the nuclear expanded analysis using sequence data from the more variable rDNA locus representing 38 individuals.* Molecular phylo- rDNA internal transcribed spacer region also reveals a phylogenetic analysis reveals two groups of species in Pleurotus, genetically based pattern of genetic divergence associated with with one ancient group of broad worldwide distribution and allopatric speciation among populations from different conti- a second group of species evolving much more recently nents in the Northern Hemisphere. The ability of rDNA within the Northern Hemisphere. These results demonstrate sequences to resolve phylogenetic relationships among geo- the utility ofrDNA phylogenies for understanding patterns of graphically isolated populations within intersterility groups evolution and speciation in basidiomycete fungi within a illustrates the importance of biogeography for understanding biogeographical context. speciation in Pleurotus. Patterns of geographic distribution among intersterilit groups suggest that several species lineages evolved quite early, with recently evolved groups re- MATERIALS AND METHODS stricted to the Northern Hemisphere and older lineages occur- Cultures and Mating Compatibility Studies. The fungal ring throughout the world. Based on phylogenetic evidence, strains used in this study are listed in Table 1 along with other analysis of historical biogeography using area cladograms data on their geographic origin and intersterility group. All shows that multiple dispersal and vicariance events are respon- strains were grown and stored on YPSS/2 agar (14). Tester sible for patterns of speciation observed. strains for determining mating compatibility relationships representing five intersterility groups were available from Speciation in many mushroom groups is often associated with previous studies (4, 15). Spore prints were obtained from tremendous levels of genetic divergence that suggest an additional strains by fruiting them on sterilized rye grain ancient origin for some species (1). Within many mushroom substrate in the laboratory. Patterns of mating compatibility species, clear patterns of morphological and genetic diver- among single-spore isolates originating from the same parent gence among geographically distinct populations also suggest (intrastock crosses) were used to determine mating system. an allopatric mode for speciation (2). The combined study of Interstock pairings among single-basidiospore isolates from phylogeny and biogeography provides a framework for un- different parent strains were used to determine the limits of derstanding the relationship among different components of additional intersterility groups. Formation of clamp connec- evolution at the species level, including geographic variation, tions in positive matings was used as the primary criterion for genetic isolation mechanisms, and morphological evolution mating compatibility (15). (3). The recent development of rigorous molecular phyloge- DNA Amplification and Sequencing. Cultures of each iso- netic approaches has made it possible to reexamine many late were grown in liquid YPSS/2 medium, and DNA mini- classic questions regarding the importance of biogeography preparations were prepared using CTAB extraction buffer as a primary factor involved in speciation. described by Zolan and Pukkila (16). PCR amplification was The oyster mushroom Pleurotus ostreatus and its related performed using reagents and primers described in Vilgalys species are among the more conspicuous fungi causing wood and Hester (17) and White et al. (19). Primers 5.8SR and LR7 decay in terrestrial ecosystems worldwide and are widely were used to amplify a 1.7- to 1.8-kb fragment of rDNA collected and cultivated as edible fungi. Mating compatibility including the internal transcribed spacer 2 (ITS-2) region and studies have demonstrated the existence of discrete interste- RNA region, and rility groups (biological species) in Pleurotus, many of which the 5' halfofthe large subunit (LSU) coding are broadly distributed over one or more continents (4). Abbreviations: indel, insertion/deletion; ITS, internal transcribed spacer; LSU, large subunit; CI, consistency index; RI, retention The publication costs ofthis article were defrayed in part by page charge index. payment. This article must therefore be hereby marked "advertisement" *The sequences reported in this paper have been deposited in the in accordance with 18 U.S.C. §1734 solely to indicate this fact. GenBank data base (accession nos. U04058-U04160). 4599 4600 Evolution: Vilgalys and Sun Proc. Natl. Acad. Sci. USA 91 (1994) Table 1. Pleurotus strains studied, with information concerning their intersterility groups, geographic origin, and other data Intersterility group Strain,* geographic origin, other culturest Comments I 261, Virginia; 403, Arizona; 850, California; 330, Includes P. ostreatus (with several varieties), presently known Czechoslovakia; 331, Czechoslovakia; 334, Germany; only from North America and northern Eurasia. Deciduous 1743, Japan (= TMI 30054); 1742, Japan (= TMI 30055) hardwood logs are the preferred substrates for this group. II 263, Virginia; 352, North Carolina; 700, British Columbia; Known in Europe as Pleurotus pulmonarius, widely distributed 479, Germany; 480, Germany; 481, Sweden; 1745, Japan across Eurasia and North America. Depending on where (= TMI 30632); 1748, Japan (= TMI 30058) they are found, group II populations occur on coniferous or deciduous logs. III 399, Maryland; 765, Montana Group III (Pleurotus populinus) is known only from North America, where it is mainly restricted to aspen species (Populus spp.), on which it is sometimes a weak pathogen (7). IV 383, England; 1166, Germany (= ATCC 38547); 1030, Collected in Europe as Pleurotus cornucopiae, but also as a Czechoslovakia (= ATCC 42045); 1738, Czechoslovakia morphologically distinct variant Pleurotus citrinopileatus in (= TMI 30141); 1736, Czechoslovakia (= TMI 30142) eastern Asia. Found on a variety of deciduous hardwood substrates. V 1847, Mexico; V, 1830, Malaysia (= ATCC 38137); V, Pantropical, occurring in warm regions all over the world. 1829, Malaysia (= ATCC 38138) Numerous species have been described in this group based on morphology, including Pleurotus diamor, Pleurotus flabellatus, Pleurotus salmoneostramineus, and Pleurotus salmonicolor (8, 9). Angiosperm wood is the preferred substrate. VI 625, Europe (= VT 1477, Penn State MW 85); 643, Europe Represents two morphologically unique species, Pleurotus (= VT 458, CBS 130.21); 1832, Czechoslovakia eryngii (with several varieties) from Europe and the Middle (= ATCC 36047); 1821, India (= ATCC 52666); 1822, East and Pleurotus fossulatus from Afghanistan. The former India (= ATCC 62885) species is restricted to the root stocks of shrubby Eryngium species on which it is known to be a weak parasite. Substrate preference of P. fossulatus is not known. VII 418, Louisiana; 420, Indiana; 478, Mexico Occurs throughout the world on angiosperm wood where it is most commonly identified as Pleurotus cystidiosus (10, 11). Geographic variants of this group are known from Taiwan as Pleurotus abalonus (12) and in Mexico as Pleurotus smithii (13). VIII 691, Pennsylvania (= Penn State MW84); 764, North Growing on angiosperms in Northern and Southern Carolina Hemispheres (North America, Europe, and New Zealand), referable to the morphological species Pleurotus dryinus. *Duke Culture Collection. tOther culture collections; ATCC, American Type Culture Collection, Rockville, MD; TMI, Tottori Mycological Institute, Tottori, Japan; CBS, Centraalbureau von Schimmelcultures, Baara, Netherlands; VT, Virginia Tech Mycological Culture Collection, Blacksburg, VA; Penn State Fungal Culture Collection, State College, PA. the amplified DNA was purified using Magic Prep columns data sets (sampled from informative characters) with random (Promega). DNA sequencing was performed using a cycle addition of sequences during each heuristic search. Evolu- sequencing reagent kit (GIBCO) with 'y33P]ATP (NEN). tionary patterns of mating behavior and biogeography were Approximately 980 bases were sequenced for 28 strains from examined using MACCLADE (25). the 5' portion of the LSU rDNA using primers LROR, LR5, LR16, and LR21. The ITS-2 region (341 bases) was se- RESULTS quenced from the same templates using primers LR15, LR1, and 5.8SR. The ITS-1 region (250 bases) was separately Mating Compatibility and Intersterility Groups. As in pre- amplified and sequenced using primers ITS1 and ITS4. vious studies with other Pleurotus species, mating reactions Molecular Evolution and Phylogenetic Anadysis. Sequences among single-basidiospore isolates from individual parental were aligned using CLUSTALV (20) and later adjusted by eye strains were consistent with a bifactorial mating incompati- using the SeqApp editor (21) to produce separate LSU RNA bility system in every case (26). Patterns of mating compat- and ITS data sets (copies of the original data sets used for all ibility among interstock pairings grouped the strains from this analyses are available from R.V. by mail or via e-mail at study into eight intersterility groups (biological species). [email protected]). Levels of molecular sequence di- Tester strains for each group have been deposited with the vergence in each data set were compared by calculating American Type Culture Collection. pairwise estimates of nucleotide substitution rates by the MolecularEvolution and Phylogeny. Sequencesfor27 strains two-parameter method of Kimura (22) using PHYLIP (23). from the 5'-most portion of the LSU rlDNA were easily Relative rate differences between the two data sets were alignable. Within the LSU data set, several small regions of compared by linear regression ofpaired distance values from uncertain aliment due to short insertions/deletions (indels) each distance matrix. Phylogenetic relationships were esti- were excluded from subsequent analyses. By contrast, align- mated for each data set using PAUP(24). Maximum parsimony ment ofall sequences was more problematic forthe ITS rDNA trees were identified uping the heuristic search option in PAUP region. Indels and other regions of uncertain nucleotide se- using only informative characters with at least 30 random quence homology were more prevalent within the ITS-i and input orders of sequences. Support for phylogenetic group- ITS-2 regions, such that it was not possible to align all of the ings was assessed by bootstrap analysis using 200 replicate most divergent taxa. Positions lacking indels forthe ITS rDNA Evolution: Vilgalys and Sun Proc. NatL. Acad. Sci. USA 91 (1994) 4601 region were used to calculate pairwise distance values among Distribution all the taxa for comparison with the LSU rDNA data set. Northern Southern Overall levels of nucleotide substitution were higher for the Hemisphere Hemisphere ITS data set (range = 0-0.41) than for the LSU sequences mI (range = 0-0.05). A linear regression of pairwise distances calculated for ITS versus LSU data sets showed a positive VI 0 linear relationship (P < 0.001), with a slope of 7.65, which provides an estimate of the difference in substitution rates between the two regions (Fig. 1). I 0 Maximum parsimony analysis ofthe LSU rDNA data with all eight intersterility groups yielded two most parsimonious trees of94 steps each, each with a consistency index (Cl) = 0.77 and retention index (RI) = 0.88 (Fig. 2; only one tree shown). II 0 Because of the lack of a satisfactory outgroup, the resulting trees were provisionally rooted along their longest internal branch by midpoint rooting. Both trees had identical topologies 420 VII 0 except for a reversal in branching order between intersterility - 478 groups VII and VIII. Bootstrap replication frequencies show VIII 0 0 most internal branches to be well supported by the LSU data, IV - 1847 V 0 with significant support for branches leading to intersterility 1829 S groups V, IV, VII, and VIII. Relationships among four remain- ing groups (I, II, III, and VI) were only partly resolved by LSU FIG. 2. Phylogenetic relationships among eight intersterility groups in Pleurotus based on LSU rDNA sequences. The unrooted data, although parsimony analysis and bootstrap values tree shown here is based on maximum parsimony analysis using strongly support their separate grouping (96%) from the remain- sequence data from the 5' portion of the LSU RNA coding region ing taxa. The relatively shallow branching ofthese four mating from 28 individuals. Significant bootstrap replication frequencies groups indicated by the phylogram in Fig. 1 suggests that they (above 50%) are indicated, with data on intersterility groups (roman may be ofrecent origin. Levels ofnucleotide substitution within numerals) and their geographic distribution. this "northern hemisphere" group are much lower (<0.01) relative to that observed between the other mating groups, several internal branches as well as nearly all terminal clades. which are believed to be of more ancient origin. Two major divisions can be recognized from the ITS phylo- Because of difficulties aligning ITS sequences across all gram, with one branch giving rise to several geographic mating groups, phylogenetic analysis of the ITS data was populations of intersterility group II and another well- restricted to a larger sampling of strains with diverse geo- supported lineage giving rise to groups I, VI, and III. Within graphic origin belonging within the northern hemisphere this second lineage, intersterility groups VI and III are groups (I, II, III, VI) along with an outgroup (VII) for rooting inferred to have evolved from within intersterility group I. purposes. Alignments for this set of sequences were not problematic and resulted in a data set with only several DISCUSSION single-base indels, which were included in the phylogenetic analysis to provide maximum resolution of phylogenetic Intersterility Groups. Many studies have employed mating relationships. Maximum parsimony analysis of the ITS data criteria to delimit intersterility groups in Pleurotus (4, 7-10, yielded a single tree with a length of 179 steps, with a CI = 15, 27-30). Together with other recent studies that used 0.89 and RI = 0.94 (Fig. 3). Phylogenetic relationships among common reference strains with ours, this study has been able all 10 population samples were completely resolved using the to identify at least eight intersterility groups of Pleurotus ITS data, with high bootstrap values providing support for (Table 1). Several intersterility groups are represented by populations from several different continents, and additional 0.5 groups remain to be identified. By employing a standard

rDNA ITS sequence 399 (97%) II N. America 591 bp 0.4 1821 Asia length = 179 steps 1822 (100%) VI (India) CI = 0.89 RI = 0.94 643 (87%) VI Europe 0 1832 0 0.3 261 1403 (76%) I N. America ._cn 850 In 330 0.2 331 (62%) I Europe 334 1742 (100%) I Asia (Japan) 1745 II 0.1 p1748 (97%) Asia (Japan) 479 480 (<50%) II Europe 0.0 ?- 1481 0.0 0.1 0.2 0.3 0.4 0.5 700 II W. N. AmericIca 263 (72%) II E. N. Amercca LSU rDNA distance i 418 100% ' 420 VII (outgroup) FIG. 1. Higher rates of substitution observed within ITS versus LSU rDNA sequence data from Pleurotus species. The plot shows FIG. 3. Maximum parsimony tree for 26 ITS sequences repre- linear regression of pairwise substitution estimates and standard senting 10 populations belonging to four intersterility groups (I, II, deviations (in parentheses) for ITS and LSU data sets [estimates III, VI) in Pleurotus. The tree is rooted using three sequences from based on the two-parameter model of Kimura as implemented by intersterility group VII. Bootstrap replication frequencies and geo- PHYLIP (23)]. graphic origin are also indicated. 4602 Evolution: Vilgalys and Sun Proc. Natl. Acad. Sci. USA 91 (1994) series of tester strains for each group, mating tests provide a convenient method for detecting new groups. Since complet- ing the molecularportion ofthis study at least three additional intersterility groups have subsequently been identified (un- published results). Because of their ephemeral fruiting patterns, the ranges and distributions of most mushroom species are poorly known. Based on what is currently known about their distributions, different intersterility groups in Pleurotus may be grouped according to whether they are primarily distributed in the Northern Hemisphere or more broadly across both hemispheres (Fig. 2). A summary of our current knowledge about the distribution ofthese groups is presented in Table 1. Biological and Phylogenetic Species. Mating criteria are widely employed by basidiomycete systematists (27), who regularly employ biological species concepts for species circumscription (2). One important observation from mating tests in basidiomycetes is that intersterility barriers between most species are prezygotic and nearly always absolute. Although few examples are known of partial intersterility or FIG. 4. Summary of phylogenetic relationships in Pleurotus hybridization in basidiomycetes, genetic analyses of infre- showing patterns of mating behavior and geographic vicariance quent crosses between reproductively isolated groups in the associated with speciation. Heterobasidion annosum complex have shown that intersterility may be controlled by as few as one or two loci (28, 29). populations within several intersterility groups. This resolu- Because mating criteria are so easily testable in saprobic tion enables phylogeographic analyses of vicariance and fungi, mushrooms may represent the best possible example of speciation. The high levels of bootstrap support for near- a group of organisms where a biological species concept is terminal branches leading to geographically defined popula- broadly applicable. Biological species concepts have always tions within each intersterility group provide strong support been controversial, however, since species grouping based for the view that speciation in these fungi is associated with on mating may not always accurately reflect evolutionary geographic isolation. Numerous studies with other mush- relationships (7, 8, 18). room groups also demonstrate clear patterns of genetic The availability of a completely resolved phylogeny at the divergence among geographically isolated populations that population level provides a unique opportunity to examine are still capable of crossing (1, 2, 10). Congruence between the phylogenetic basis for patterns of mating behavior and patterns of rDNA evolution with mating behavior and bio- vicariance in Pleurotus. A cladogram summarizing patterns geography suggests that processes that might otherwise con- of biogeography and mating relationships shows that most found analysis of organismal phylogeny (e.g., paralogy, in- lineages resolved by molecular phylogenetic analysis (groups terspecific gene flow) are not likely to be a significant factor II, III, V, IV, VII, and VIII) are also congruent with species for analysis of speciation at the population levels sampled in groupings based on a biological species concept (Figs. 2-4). this study. Within the more recent northern hemisphere group, how- The phylogenetic relationships among different popula- ever, phylogenetic analysis shows two intersterility groups to tions also provide a framework for understanding the bio- be paraphyletic, with different geographic populations of geographic history of speciation in Pleurotus. Geographic intersterility group I giving rise to populations of group VI, distribution patterns show the earlier evolving species groups which itself is paraphyletic with respect to mating group III (IV, V, VII, VIII) to be broadly distributed across Northern (Figs. 3 and 4). As pointed out by Cracraft (8), many broadly and Southern Hemispheres, while the more recently evolved distributed biological species are often found to be paraphyl- groups (I, II, III, VI) appear to be largely restricted within the etic and thus may not always represent unique evolutionary Northern Hemisphere. Two possible historical explanations units. This may also be the case in Pleurotus as well. are proposed to account for this distribution. One explana- However, resolution of these paraphyletic relationships is tion, based on continental drift theory, is that the early not strongly supported by either bootstrap analysis or par- species groups were already present before the breakup of simony (an alternative tree in which all intersterility groups the Pangean continent 200 million years ago, with subsequent are monophyletic is only three steps longer than that shown "recent" speciation events occurring in the Northern Hemi- in Fig. 3). Based on the possibility that some intersterility sphere with the breakup of Laurasia (100 million years ago). groups may in fact be paraphyletic, we and others have This scenario, however, predicts an extremely ancient origin proposed that phylogenetically based species concepts (in for Pleurotus and other groups ofbasidiomycetes that would which monophyletic geographic populations could represent be predicted to share similar patterns of gross vicariance species groups) should be more desirable for developing across the Northern and Southern Hemispheres. Under this species concepts in mushrooms and other fungi (2). The hypothesis, origins ofhigher fungal taxa would probably need largely concordant groupings based on molecular phylogeny to be dated back to the Cambrian or even earlier. An with groupings based on intersterility, however, nevertheless alternative hypothesis that does not require such drastically still also provide a strong phylogenetic basis for continued ancient origins is based on intermittent but regular dispersal recognition of biological species as fundamental taxonomic of species over broad distances to new areas. Under such a units (9). diffusion model, geographic distributions ofspecies would be Biogeography and Speciation. Cladistic analysis of LSU correlated with ages since each species diverged. Older and ITS sequences provides a completely resolved phylog- species lineages would therefore be predicted to have broader eny for intersterility groups in Pleurotus, with robust support distributions, while recent species groups would have more for most internal and terminal clades. Most significantly, the narrow distributions. 7-fold higher rate of substitution in the ITS sequences than in Phylogenetic evidence also provides a framework for ex- the LSU sequences provided sufficient variation for resolv- amining patterns of speciation based on vicariance biogeog- ing phylogenetic relationships among geographically isolated raphy. Within the Northern Hemisphere, at least two and Evolution: ViIgalys and Sun Proc. Natl. Acad. Sci. USA 91 (1994) 4603 possibly three separate North American-Eurasian vicariance 8. Cracraft, J. (1990) in Speciation and Its Consequences, eds. events (along with dispersal) are required to account for the Otte, D. & Endler, J. A. (Sinauer, Sunderland, MA), pp. 28-59. 9. Avise, J. C. & Ball, R. M., Jr. (1990) Oxford Surveys Evol. phylogenetic hypothesis presented in Fig. 4. Two contrasting Biol. 7, 45-67. biogeographic histories are necessary to explain phylogenetic 10. Garbelotto, M., Bruns, T. D., Cobb, F. W. & Otrosina, W. J. history in intersterility groups I and II, which differ in their (1993) Can. J. Bot. 71, 565-569. interpretation of relationships among European, North 11. Pegler, D. N. (1977) Kew Bull. 31, 501-510. American, and Asian populations. The evolution of two 12. Pegler, D. N. (1986) Agaric Flora ofSri Lanka (Her Majesty's Stationary Office, London). biogeographically diverse intersterility groups from within 13. Moore, R. T. (1985) Trans. Br. Mycol. Soc. 85, 354-358. group I also suggests evidence for several recent species 14. Vilgalys, R., Smith, A., Sun, B. L. & Miller, 0. K., Jr. (1993) radiations that have occurred within the Northern Hemi- Can. J. Bot. 71, 113-128. sphere, possibly associated with Pleistocene glaciation. 15. Eger, G. (1978) in The Biology and Cultivation of Edible Because the meager fossil record for fungi provides little Mushrooms, eds. Chang, S. T. & Hayes, W. A. (Academic, data on the origin and dispersal of species, molecular phy- New York), pp. 497-519. 16. Zolan, M. E. & Pukkila, P. J. (1986) Mol. Cell. Biol. 6,195-200. logenetic studies may be the only means for reconstructing 17. Vilgalys, R. & Hester, M. (1990) J. Bacteriol. 172, 4238-4246. evolutionary history and for testing biogeographic hypothe- 18. Donoghue, M. J. (1985) The Bryologist 88, 172-181. ses. The ability of rDNA sequence data to provide complete 19. White, T. J., Bruns, T., Lee, S. & Taylor, J. W. (1990) in PCR phylogenetic resolution at the population level provides a Protocols: A Guide to Methods and Applications, eds. Innis, window through which mycologists might view the ancient M. A., Gelfand, D. H., Sninsky, J. J. & White, T. J. (Academ- and recent origins of fungal species within a biogeographic ic, New York), pp. 315-322. context. 20. Higgins, D. G. & Sharp, P. M. (1989) Comput. Appl. Biol. Sci. 5, 151-153. 21. Gilbert, D. G. (1992) SEQAPP (Indiana University, Blooming- We thank T. D. Bruns, J. J. Doyle, and B. G. Baldwin for their ton), Version 1.9a162. helpful comments. This work was supported by National Science 22. Kimura, M. (1980) J. Mol. Evol. 16, 111-120. Foundation Grant DEB 91-07812. 23. Felsenstein, J. (1993) PHYLIP (University of Washington, Se- attle), Version 3.5. 1. Vilgalys, R. J. & Johnson, J. L. (1987) Proc. Natl. Acad. Sci. 24. Swofford, D. L. (1991) PAUP, Phylogenetic Analysis Using USA 84, 2355-2358. Parsimony (Ill. Nat. Hist. Surv., Champaign, IL), Version 2. Vilgalys, R. (1991) Mycologia 83, 758-773. 3.1.1. 3. Avise, J. C. (1989) Evolution 43, 1192-1208. 25. Maddison, W. P. & Maddison, D. R. (1991) MACCLADE, Inter- 4. Anderson, J. B. & Stasovski, E. (1992) Mycologia 84,505-516. active Analysis of Phylogeny and Character Evolution (Sin- 5. Moncalvo, J. M., Rehner, S. A. & Vilgalys, R. (1993) Myco- auer, Sunderland, MA), Version 3.0.3. logia 85, 788-794. 26. Hilber, 0. (1982) Bibl. Mycol. 87, 1-448. 6. Zambino, P. J. & Szabo, L. J. (1993) Mycologia 85, 401-414. 27. Boidin, J. (1986) Mycotaxon 26, 319-336. 7. Mayr, E. (1982) The Growth ofBiological Thought (Belknap, 28. Chase, T. E. & Ullrich, R. C. (1990) Mycologia 82, 73-81. Cambridge, MA). 29. Chase, T. E. & Ullrich, R. C. (1990) Mycologia 82, 67-72.